USDT(TRC20): TAaBpuhdoPGNv9xnyne2nXQ5gLQLNUrY6V
2. Image upscaler
3. NSFW Detector
4. Image Analyzer
5. Video Analyzer
6. Prompt Generator
7. [BETA]AI Detector
1. Windows 10-11
2. NVIDIA GPU with CUDA 11.8 support
3. 16gb RAM
4. 8gb+ free SSD (NOT HDD) space
1. Install winget
2. Install Python
winget install -e --id Python.Python.3.10
3. Install Git
winget install -e --id Git.Git
4. Install FFmpeg
winget install -e --id Gyan.FFmpeg
5. Install CUDA 11.8
6. Restart your PC
7. Clone repository
git clone https://github.com/ShamHyper/MultiAI.git
8. Run install.bat for first time. Next time run run.bat
1. Install Python 3.10.x
2. Install Git
3. Install FFmpeg
4. Install CUDA 11.8
5. Restart your PC
6. Download using "Code --> Download ZIP" (or click here)
7. Unarchive ZIP
8. Run install.bat for first time. Next time run run.bat
1. Install Python
sudo apt install python3
2. Install PIP
sudo apt install python3-pip
3. Install Git
sudo apt install git
4. Install FFmpeg
sudo apt install FFmpeg
5. Install CUDA 11.8
6. Restart your PC
7. Clone repository
git clone https://github.com/ShamHyper/MultiAI.git
8. In terminal, type for first time:
pip3 install -r requirements.txt
9. In terminal, type for run MultiAI:
python3 app.py
This is a very BETA version of the installation for Linux. I have briefly checked the functionality and do not really understand how to set up a venv for this OS yet.
BgRemoverLite | Image upscaler | NSFW Detector | Image Analyzer | Video Analyzer | Prompt Generator | AI Detector |
---|---|---|---|---|---|---|
Fully functional | Partially functional | Not tested | Not tested | Not tested | Not tested | Not tested |
Now you don't have to go into config.json and change something with your hands. This tab will do everything for you!
- Choose necessary checkboxes
- Press Save settings
- Restart MultiAI
{
"debug_mode": "True",
// Boolean ("True" or "False")
// Enable debug mode (write debug info)
// Enable logging
"start_in_browser": "True",
// Boolean ("True" or "False")
// Enable MultiAI starting in browser
"share_gradio": "False",
// Boolean ("True" or "False")
// Enable MultiAI starting with share link
"preload_models": "False",
// Boolean ("True" or "False")
// Preloading AI models
// False loading time: ~5-10s
// True loading time: ~40-60s
"clear_on_start": "False"
// Boolean ("True" or "False")
// Clear all outputs on start
}
- If you see errors like this:
ModuleNotFoundError: No module named "certifi"
, run install.bat for fix them. If that didn't help, open a new issue - If you encounter errors during the execution of the program, open a new issue
- If your console window freezes during the install process.bat, restart it with administrator rights. It is also highly recommended to run run.bat from the administrator.
1. BgRemoverLite 2. Image upscaler 3. NSFW Detector 4. Image Analyzer 5. Video Analyzer 6. Prompt Generator 7. [BETA]AI Detector
So far, most of the libraries used in my project have not been updated to 3.11 - 3.12 or are working incorrectly. I tried to adapt the project for the new version, but for now I advise you to stay within 3.10. After v1.10.x, discontinued support for 3.9, only 3.10 now
- Built with Gradio - https://www.gradio.app/
- Using u2net model for BgRemoverLite - https://github.com/xuebinqin/U-2-Net
- RemBG library for BgRemoverLite - https://github.com/danielgatis/rembg
- NSFW Detection - https://github.com/GantMan/nsfw_model
- Upscaling - https://github.com/kmewhort/upscalers
- Clip-interrogator for Image Analyzer - https://github.com/pharmapsychotic/clip-interrogator
- Fast Prompt Generator - https://huggingface.co/FredZhang7/anime-anything-promptgen-v2
- H5 Keras Model - https://github.com/Shahnawax/HAR-CNN-Keras/
MIT license | Copyright (c) 2023-2024 ShamHyper